SciTransfer
SOLARIS · Project

AI Deepfake Detection and Trust Analysis for Digital Content Security

digitalTestedTRL 4

Imagine a world where you can't tell if a video of a politician or CEO is real or a computer-generated fake. This work figures out why our brains trust these fakes and how to spot them. It also explores how to use this same AI tech to create positive, helpful messages instead of lies.

By the numbers
13
partners
8
countries
23%
industry ratio
The business problem

What needed solving

Companies and governments struggle to distinguish AI-generated deepfakes from reality, leading to disinformation and trust erosion. There is a lack of technical and psychological tools to mitigate these risks in real-time.

The solution

What was built

A psychometric scale of perceived trustworthiness for deepfakes and a set of simulation protocols for viral deepfake spreading.

Audience

Who needs this

Cybersecurity firmsGovernment communication ministriesDigital news platformsSocial media moderation teams
Business applications

Who can put this to work

Cybersecurity
mid-size
Target: Digital identity verification firm

If you are a verification firm dealing with identity theft via AI-generated video — this project developed a psychometric scale of perceived trustworthiness that helps identify why users believe fakes, allowing for better detection tools.

Media & Journalism
enterprise
Target: News agency

If you are a news agency dealing with viral disinformation — this project developed simulation protocols for the spreading of deepfakes that help organizations create mitigation strategies to protect their reputation.

Public Relations
SME
Target: Social impact marketing agency

If you are a marketing agency dealing with low citizen engagement on global issues — this project developed value-based GAN contents that use AI to create inclusive and constructive messages for the public.

Frequently asked

Quick answers

What is the cost or price for implementing these tools?

Based on available project data, there is no specific pricing or commercial cost listed for the resulting tools.

Is this technology ready for industrial scale?

The project uses use-cases and simulations to test protocols, but based on available project data, it has not yet reached full industrial scale.

How is the IP or licensing handled?

Based on available project data, specific licensing terms or patent details are not provided.

What regulations does this address?

The project focuses on establishing regulatory innovations to detect and mitigate deepfake risks to protect democratic governance.

What is the timeline for deployment?

The project period runs from 2023-02-01 to 2026-01-31.

Consortium

Who built it

The consortium is heavily weighted toward academic research with 7 universities and 2 research institutes. However, there is a significant business presence with 3 industry partners, including 2 SMEs, representing a 23% industry ratio. This suggests the project is primarily research-driven but has a clear path toward commercial application through its industry members across 8 countries.

How to reach the team

Universiteit Utrecht

Next steps

Talk to the team behind this work.

Contact us to connect with the SOLARIS consortium for deepfake mitigation licensing.